Building structure discovery and display from various data artifacts at scene

BACKGROUND

The invention pertains to buildings and particularly to their structures. The invention more particularly pertains to determining structural components of a building.

SUMMARY

The invention is an approach for developing information about a building of which little or no information is known at a time when firefighters need to enter the building. The approach may incorporate obtaining an image of the building from a publicly available resource. An outside boundary of the building may be determined from the image. The number of floors may be determined. A 3D floor plan may be generated from the boundary and the number of floors. Structural components may be added to the 3D floor plan, which are deduced from movements of firefighters, and the movements may be deduced from their locations. The locations may be transmitted by devices on the firefighters. Structural components of the building indicated by persons in the vicinity of the building may also be added to the 3D floor plan.

BRIEF DESCRIPTION OF THE DRAWING

<figref idref="DRAWINGS">FIG. 1</figref> is a diagram of a satellite image of a building structure for which information is to be discovered;

<figref idref="DRAWINGS">FIG. 2</figref> is a diagram of the satellite image with an outline of the building structure delineated;

<figref idref="DRAWINGS">FIG. 3</figref> is a diagram of a generated 3D floor plan of the building structure;

<figref idref="DRAWINGS">FIG. 4</figref> is a diagram of the generated 3D floor plan with a structure component added;

<figref idref="DRAWINGS">FIG. 5</figref> is a graph revealing a motion of a person in a building structure from three different axes;

<figref idref="DRAWINGS">FIG. 6</figref> is a diagram of the motion of a person as locations shown on a 2D floor plan;

<figref idref="DRAWINGS">FIG. 7</figref> is a diagram of the 3D floor plan with a symbol entered representing a structure component;

<figref idref="DRAWINGS">FIG. 8</figref> is a diagram of the 2D floor plan indicating an identification of a hallway or path with locations having some distribution entered as symbols;

<figref idref="DRAWINGS">FIG. 9</figref> is a 3D floor plan having symbols that represent objects and associated behaviors; and

<figref idref="DRAWINGS">FIG. 10</figref> is a diagram indicating an inference of structural components from data fusion.

DESCRIPTION

While some modern buildings provide digital or at least paper floor plans/maps to local fire departments, many legacy buildings still do not. Firefighters arriving on the scene may be forced to assess a fire situation with little or no information about the interior building structure except perhaps from what they might recall from a previous experience at the building. What is needed is a way to automatically derive an approximation of the interior structure from data artifacts available at the scene.

The present approach is to first automatically derive the exterior structure of a building as a wireframe or 3D floor plan using a shape extraction processing on public source images of the building, such as Google Earth™ (Google Earth). If a public data source also includes a perspective view image of the building, then one may discover from windows and other features of the building, the number and positioning of the interior floors of the building. Going further, one may then discover the approximate locations of various interior building structures such as stairways, elevators, hallways and other types of spaces from the locations and movements of firefighters reported by a location tracking system, and fire alarm system data.

One may track a firefighter with an IMU (inertial measurement unit) via dead reckoning, RSSI (received signal strength index), or the like. Now, for example, if there is no sign of body movement (from the IMU on the firefighter), but the firefighter moves from one floor to another, then one may assume an elevator at that location. If the IMU indicates movement and the firefighter moves from one floor to another, then the firefighter may be at a stairway. Straight line movement is probably a hallway. A short digression off the path to the side may indicate a room.

From addressable smoke and heat sensor data, one may obtain the location and identity of building spaces that are in alarm. One may add these items to the emerging knowledge of the interior structure.

One may then draw these internal structures that have been discovered onto the floors and the 3D exterior wireframe boundary of the building derived from the shape extraction. One may automatically insert the locations of activated smoke and heat detectors into the emerging 3D graphic of the building, as well as any other artifacts of interest to the incident commander that can be detected or reported.

One may note that in this concept, the picture of the building evolves during the incident as more tracks are reported by the tracking system and more alarms go off.

One may be able to generate 3D maps in real-time from data collected on-scene. For instance, if one is the incident commander arriving at the scene and has no building plan available, then one needs to be able to generate a 3D building plan, even if it is crude or coarse. There may be only several ways to do this. The “data collected on site” may be: 1) a sketch of the building in that the incident commander might try to do quickly, perhaps on a digitizer pad; 2) One might take photographs or video of the building; or 3) One might pull down the Google Earth images of the building (very coarse in resolution).

From the photos or Google Earth images, one may extract some features of the building exterior. One may even infer how many floors the building has from the placement of the windows. One may capture the main entrances. One may use the scale of the image to estimate building dimensions. From a drawing on a digitizer pad, one may do similar things. Thus, one may draw up a shell of a building populated with a certain number of blank floors.

One may generate the building exterior from a sketch on the digitizer pad, photos and a Google Earth image. As to an inside of a building, one may infer certain things from tracking data. If there is no sign of body movement (according to the IMU on the firefighter) and the firefighter can move from one floor to another, then one may assume an elevator to be in the building. If the IMU indicates movement and the firefighter moves from one floor to another, then a stairway may be assumed. Straight line movement may imply a hallway. A short digression off the path to the side may imply a room.

One may deduce more information (inside the structure) via a fusing of data (body movement from an IMU plus body location). One may then draw an internal structure on the exterior boundary.

One may track the firefighter with an IMU (inertial measurement unit) via dead reckoning. One may also track the firefighter with an RSSI (received signal strength index) via a nearest neighbor lookup (NNL), and other items. Some of these items may have disadvantages. The IMU appears to suffer from accumulated error and the RSSI appears to suffer from a dynamic environment, such as temperature and humidity, and so on. However, combining these items may result in a tracking system.

Additionally, “data collected on site” may also incorporate a firefighter's location, triggered alarms, and so on. So in this case, the 3D map may overlay these data on the building plan in real time. It is important to let the first command browse numerous information overlays on the 3D map.

<figref idref="DRAWINGS">FIG. 1</figref> is a diagram of a satellite image <b>11</b> of the building structure to be discovered and displayed from various data artifacts. The longitude and latitude <b>13</b> may be obtained from a GPS (global positioning system) device. The satellite image <b>11</b> may be displayed and saved. A symbol <b>12</b> representing a truck of the first responder is shown on the image <b>11</b>. The image may be panned and zoomed in and out with buttons <b>14</b> to get the destination which may be a building structure <b>15</b>.

<figref idref="DRAWINGS">FIG. 2</figref> is a diagram of satellite image <b>11</b> of structure <b>15</b> with a structure area selected with a dashed line rectangle <b>16</b>, using a button <b>17</b> labeled “Set Area”. An image process module may refine the boundary of building <b>15</b> of <figref idref="DRAWINGS">FIG. 2</figref> with a color or a boundary <b>18</b>. A “Set O Point” button <b>19</b> may be for setting an origin point <b>21</b> of a tracking system.

<figref idref="DRAWINGS">FIG. 3</figref> is a diagram showing a generation of a 3D floor plan from structure <b>15</b> of <figref idref="DRAWINGS">FIG. 2</figref>. A “spine button” <b>22</b> may be pressed to set a number of floors. Button <b>22</b> may be pressed to show “2”. With boundary <b>18</b> and the number of floors being two, the 3D model, floor plan or wireframe <b>23</b> of structure <b>15</b> may be generated and displayed.

<figref idref="DRAWINGS">FIG. 4</figref> is a diagram of the 3D floor plan <b>23</b> from <figref idref="DRAWINGS">FIG. 3</figref> but shows an adding of a structure shown with a symbol <b>24</b> on a first floor <b>51</b>. For example, when a stair is identified, the 3D model <b>23</b> may be updated by adding the stair represented by symbol <b>24</b>. When an elevator is identified, 3D model <b>23</b> may likewise be updated with a symbol <b>25</b> representing the elevator on a second floor <b>52</b>. Other structures and items of building <b>15</b> may similarly be represented with a placement of symbols in appropriate places on the 3D model <b>23</b> floor plan.

<figref idref="DRAWINGS">FIG. 5</figref> is a graph <b>26</b> showing the motion of, for example, a person climbing from views of the X, Y and Z axes. The graph is in terms of value (distance) versus time. The X axis movement is shown by plot <b>27</b> which may represent motion on a stairway <b>30</b> represented in a 2D floor plan in <figref idref="DRAWINGS">FIG. 6</figref> in the direction of an arrow <b>31</b>. The Y axis movement is shown by plot <b>28</b> which may represent motion on stairway in the direction of an arrow <b>32</b>. The Z axis movement is shown by plot <b>29</b> which presents motion on the stairway <b>30</b> in the direction of arrow <b>33</b> coming out of the diagram at 90 degrees relative to the surface of the diagram. The dots <b>34</b> may represent positions of, for example, a person climbing the stairs <b>30</b>.

<figref idref="DRAWINGS">FIG. 7</figref> is a diagram of the 3D floor plan <b>23</b> showing an introduction on floor <b>51</b> of a symbol representing a hallway or hallways <b>38</b> once inferred or discovered as indicated in <figref idref="DRAWINGS">FIGS. 8 and 10</figref>.

Features of the structure of a building may be had from behavior definition and identification. Features of possible behaviors that can be identified may incorporate climbing stairs, as noted herein, turning left and/or right 90 degrees, travel in an elevator via acceleration, walking in a room, and so forth. <figref idref="DRAWINGS">FIG. 8</figref> shows positions, represented by dots <b>37</b>, of a person reflecting a behavior, which may indicate an existence of a hallway <b>39</b> with a 90 degree corner. Various behaviors with different features from the trajectory of personnel may be detected. Building structures may be deduced from the behaviors. A Markov-Model may be applied to analyze the time series data and then identify a behavior.

When a hallway or path is identified, the 3D floor plan <b>23</b> may be updated with the hallway or path <b>39</b> as shown in <figref idref="DRAWINGS">FIG. 8</figref>. An approach for identifying a hallway or path in a floor is shown. The dots or locations <b>37</b> may be put on an image or floor plan with a boundary, i.e., each location has some distribution (e.g., Gaussian distribution). The image or floor plan may be analyzed to get the hallway or path into the plan. The distribution of each location may be added. The image or plan may be normalized. Then some value beyond a certain threshold may be identified as indicating a hallway or a path.

An object and its behaviors may be added to a floor plan of the 3D floor plan <b>23</b> in <figref idref="DRAWINGS">FIG. 9</figref>. For instance, an object may be added to its corresponding location. Various behaviors or roles may be represented with a distinct style such as a person walking <b>41</b>, a person down <b>42</b>, a lead person or a member, and so on. These items of information may be deduced from tracking data.

<figref idref="DRAWINGS">FIG. 10</figref> is a diagram showing that more high level information may be integrated from data fusion. Data fusion may incorporate trajectories of a person at different times, trajectories of different personnel, and data from different sensors (e.g., RSSI or inertia measurement). Fusion strategies may incorporate an intersection when the data are similar, and a union when the data are anisotropic. Multiple trajectories may output information such as identifying rooms. For instance, personnel of locations <b>44</b> and personnel of locations <b>45</b> may identify or result in detection of a room or space <b>46</b>. Personnel reflected by other locations may identify or result in detection of hallways <b>47</b>, <b>48</b> and <b>49</b>.

Relevant patent documents may include U.S. Pat. No. 7,683,793, issued Mar. 23, 2010, and entitled “Time-Dependent Classification and Signaling of Evacuation Route Safety”; U.S. patent application Ser. No. 12/200,158, filed Aug. 28, 2008, and entitled “Method of Route Retrieval”; and U.S. patent application Ser. No. 12/573,398, filed Oct. 5, 2009, and entitled “Location Enhancement System and Method Based on Topology Constraints”. U.S. Pat. No. 7,683,793, issued Mar. 23, 2010, is hereby incorporated by reference. U.S. patent application Ser. No. 12/200,158, filed Aug. 28, 2008, is hereby incorporated by reference. U.S. patent application Ser. No. 12/573,398, filed Oct. 5, 2009, is hereby incorporated by reference.

In the present specification, some of the matter may be of a hypothetical or prophetic nature although stated in another manner or tense.

Although the present system has been described with respect to at least one illustrative example, many variations and modifications will become apparent to those skilled in the art upon reading the specification. It is therefore the intention that the appended claims be interpreted as broadly as possible in view of the prior art to include all such variations and modifications.